US 7738599 B2 Abstract In a quadrature amplitude (QAM) demodulator, an auxiliary symbol may be utilized in place of the decision symbol to adjust the decision-feedback loops within the demodulator. For the formation and definition of the auxiliary symbol, the radius and angle information of the received signal or of the preliminary symbol may be used. Through use of the auxiliary symbol instead of the decision symbol, any error in the angle information due to the unknown frequency and phase deviation of the local oscillator may be ignored. An auxiliary symbol generator may be provided which, instead of assigning to the received signal an element from the predetermined symbol alphabet, generates an auxiliary symbol that lies on the most probable one of the nominal radii. Nominal radii may mean those radii on which in QAM the predetermined symbols of the alphabet lie in the plane determined by the quadrature signal pair. For the angle component of the auxiliary symbol, the angle information of the sampled digital signal may be used. In polar coordinates, the auxiliary symbol may thus correspond to the vector intersection point of the sampled digital signal with the most probable nominal radius.
Claims(18) 1. A method of generating an auxiliary symbol when a digital signal locked to a quadrature signal pair is received, the method comprising:
providing nominal radii and range limits according to predetermined positions of the digital signal in a plane determined by the quadrature signal pair;
determining a preliminary symbol from the digital signal by sampling the digital signal as controlled by a symbol sampling clock;
determining polar coordinates of the preliminary symbol;
determining a nominal radius from the polar coordinates of the preliminary symbol according to the range limits, where the determined nominal radius and an angle component define polar coordinates of the auxiliary symbol in the plane of the quadrature signal pair;
selecting one of the auxiliary symbol and the preliminary symbol to provide a selected symbol; and
adjusting at least one decision-feedback controller of a demodulator in response to the selected symbol.
2. The method of
3. The method of
4. The method of
5. The method of
6. The method of
7. The method of
8. The method of
9. The method of
10. The method of
11. A circuit for generating an auxiliary symbol from a preliminary symbol in a device for receiving a digital signal locked to a quadrature signal pair, comprising:
a resolver that converts Cartesian quadrature signal components of the preliminary symbol into polar coordinates;
a radius decision stage that determines from the polar coordinates of the preliminary symbol the most probable nominal radius, where the most probable nominal radius and an angle component of the preliminary symbol define polar coordinates of the auxiliary symbol; and
a multiplexor that receives the preliminary symbol and the auxiliary symbol, and provides a selected symbol selected from the preliminary symbol and the auxiliary symbol; and
a control unit that adjusts at least one decision-feedback controller of a demodulator in response to the selected symbol.
12. The circuit of
13. A circuit for generating an auxiliary symbol from a preliminary symbol in a device for receiving a digital signal locked to a quadrature signal pair, comprising:
a resolver that converts Cartesian quadrature signal components of the preliminary symbol into polar coordinates;
a radius decision stage that determines from the polar coordinates of the preliminary symbol the most probable nominal radius, where the most probable nominal radius and an angle component of the preliminary symbol define polar coordinates of the auxiliary symbol; and
a second resolver that converts the polar coordinates of the auxiliary symbol to Cartesian coordinates in a plane determined by the quadrature signal pair;
a multiplexer that receives the auxiliary symbol and the preliminary symbol, and provides a selected signal therefrom to the at least one decision-feedback controller for control thereof,
where at least one decision-feedback controller in the device utilizes the selected signal for control thereof.
14. The circuit of
15. A method for adjusting at least one decision-feedback controller within a demodulator using an auxiliary symbol in place of a decision symbol, the method comprising:
receiving a digital signal locked to a quadrature signal pair;
determining nominal radii and range limits according to predetermined positions of the digital signal in a plane determined by the quadrature signal pair;
determining a preliminary symbol from the digital signal;
determining the auxiliary symbol from the preliminary symbol;
selecting one of the auxiliary symbol and the preliminary symbol to provide a selected symbol; and
adjusting the at least one decision-feedback controller in dependence on the selected symbol.
16. The method of
determining polar coordinates of the preliminary symbol;
determining a nominal radius from the polar coordinates of the preliminary symbol in accordance with the range limits, the determined nominal radius comprising one of the nominal radii; and
determining the auxiliary symbol in terms of polar coordinates thereof, the polar coordinates of the determined auxiliary symbol comprising the determined nominal radius and an angle component of the preliminary symbol.
17. The method of
18. The method of
Description This application claims priority from International application PCT/EP03/11099, filed Oct. 8, 2003 and German application 102 49 492.4, filed Oct. 24, 2002. This invention relates in general to digital signal processing and in particular to generating an auxiliary symbol in place of a decision symbol for adjusting a QAM demodulator as a part of a receiver in which the decision-feedback loops are not yet synchronized. Decision-feedback loops utilized in quadrature amplitude modulation (QAM) receivers typically need to be quickly brought into synchronization or “lock” when digital signals locked to a quadrature signal pair are received. Such loops are used, for example, for the adjustment of sampling instants, for the adjustment of an equalizer that removes linear distortion during the reception of the quadrature signal pair, or in an automatic gain control circuit to adapt the received signals to the dynamic range. In encoded form, these digital signals, which may also be referred to as symbols, may represent a single-bit or multiple-bit binary value. Encoding for transmission may be accomplished via the quadrature signal pair, which corresponds to a vector that at given instants of time takes up discrete positions in the amplitude and phase space of the quadrature signal pair. These instants of time typically follow each other at equal intervals and generally are sampled by the sampling clock pulses as precisely as possible. Besides QAM, another typical transmission method is phase-shift keying (PSK). In a conventional receiver for receiving digital signals, a complex multiplier or mixer, which may be controlled by a local oscillator, may downconvert the received QAM signal, which may be modulated onto a carrier frequency for transmission, to the baseband frequency. If digital signal processing is used, this downconversion can take place prior to or after analog-to-digital conversion, with the signal advantageously being sampled and digitized at the symbol rate or a multiple thereof. If the digitization rate is an even-numbered multiple of the symbol rate, each of the symbol clock pulses typically coincides with a sample value. The digitization rate may advantageously be locked to the recovered symbol rate via a phase-locked loop. Instead, if the digitization rate is free running in relation to the symbol rate, the symbol may be formed as time information via an all-digital sample-rate conversion. In this manner, a temporal interpolation between the digitized sample values of the received digital signal may be controlled. Automatic gain control circuits help to achieve a relatively high utilization of the respective dynamic range and to map the received symbols onto the symbol decision stage. An adaptive equalizer typically reduces intersymbol interference, which may result from linear distortion caused by the transmitter, the transmission path, or the receiver. In prior art demodulators for QAM or PSK signals, the circuits for controlling the frequency and phase of the local oscillator (e.g., the automatic gain control, the symbol clock recovery, and the adaptive equalizer) typically look at the differences between the received symbol and that element of the predetermined symbol alphabet which may be regarded by a decision stage as the most probable symbol that matches the received symbol. This type of control over the decision symbol is usually referred to as decision-feedback control. Since in prior-art digital demodulators the decision-feedback loops are coupled together, bringing these loops into a synchronization or lock condition may be difficult to achieve in a relatively rapid timeframe as long as the control for the carrier of the local oscillator is not yet stable in frequency and phase. Frequently, the synchronization or lock condition of the decision-feedback loops can be achieved if the respective frequencies and phases are relatively close to their desired values. Examples of decision-feedback loops are found in a book by K. D. Kammeyer, “Nachrichtenübertragung”, published by B. G. Teubner, Stuttgart, 2nd edition, 1996, pages 429 to 433, in Chapter 5.7.3, “Adaptiver Entzerrer mit quantisierter Rückführung”, pages 200 to 202, in Chapter 5.8.3, “Entscheidungsrückgekoppelte Taktregelung”, pages 213 to 215, and in Chapter 12.2.2, “Entscheidungsrückgekoppelte Trägerphasenregelung im Basisband”, pages 429 to 431. What is needed is a QAM demodulator that utilizes a relatively more reliable auxiliary symbol instead of a relatively less reliable decision symbol to adjust the decision-feedback loops within the demodulator. In a QAM demodulator, an auxiliary symbol may be utilized in place of the decision symbol to adjust the decision-feedback loops within the demodulator. For the formation and definition of the auxiliary symbol, the radius and angle information of the received signal or of the preliminary symbol may be used. Through use of the auxiliary symbol instead of the decision symbol, any error in the angle information due to the unknown frequency and phase deviation of the local oscillator may be ignored. An auxiliary symbol generator may be provided which, instead of assigning to the received signal an element from the predetermined symbol alphabet, generates an auxiliary symbol that lies on the most probable one of the nominal radii. The term nominal radii as used herein may mean those radii on which in QAM the predetermined symbols of the alphabet lie in the plane determined by the quadrature signal pair. For the angle component of the auxiliary symbol, the angle information of the sampled digital signal may be used. In polar coordinates, the auxiliary symbol may thus correspond to the vector intersection point of the sampled digital signal with the most probable nominal radius. The decision as to which nominal radius may be the most probable may be made via range limits which for example may be determined by the possible radii of the respective QAM standard, in particular by defining limit radii. These limit radii may form annuli of different widths in the quadrature signal plane which may contain one nominal radius each. It is also possible for the range limits to be determined not only by the nominal radii but also by the positions of these elements in the quadrature signal plane. In that case, the range limits may no longer define annuli but may distort the annuli. This may mean, however, that the respective angle information may influence the auxiliary symbol decision, but with little weight. Furthermore, entire regions of the quadrature signal plane can be excluded from the auxiliary decision (i.e., “masked out”) because their evaluation may be uncertain. As discussed herein, a determination of where the individual nominal radii and range limits lie may be made, so that the most probable nominal radius can be selected. Where the auxiliary symbol decision may be made via the most probable nominal radius using annuli, the radii limits may be determined, which advantageously may lie midway between two adjacent nominal radii. The respective radii or range limits may be retrieved from a table or may be continuously recalculated in accordance with the transmission standard. In higher-order QAM, some of these annuli may be so narrow that their evaluation in the presence of usual interference may be uncertain. However, since the contribution of these annuli to the control process may be relatively small, this uncertainty may be of little or no consequence. Nevertheless, the effect of the uncertain annuli can be further reduced by suitable weighting of the control information, or these annuli may be masked out. Furthermore, annuli can be permitted which enclose the respective nominal radius more narrowly and thus cover it with relatively greater certainty. If the measured radius lies outside these narrower radii limits, no auxiliary symbol may be defined, due to the relative uncertainty. For a received digital signal with the quadrature components I=R cos α and Q=R sin α that falls into an annulus of nominal radius Rsi, an auxiliary-symbol generator may form, at the position with the nominal radius Rsi and the angle α, an auxiliary symbol with the polar coordinates Rsi, α. For the auxiliary symbol to be used by the decision-feedback loops of the clock recovery, gain control, and/or equalizer, the quadrature components I The radius R and the angle α of the auxiliary symbol may be determined mathematically from the quadrature components I, Q as follows:
There are also resolvers which may convert from Cartesian coordinates to polar coordinates using other methods. In the digital signal processing portion of such resolvers, the Cordic technique may be employed, as it uses binary additions and multiplications which can be implemented by simple arithmetic shifts. Furthermore, other approximation methods or tables are possible. For the inverse conversion, i.e., for the conversion from polar signal components R and α to their quadrature components I=R cos α and Q=R sin α, a Cordic converter, a table, or an approximation method can be used. These and other objects, features and advantages of the present invention will become more apparent in light of the following detailed description of preferred embodiments thereof, as illustrated in the accompanying drawings. Referring to The graphs of However, a different result may be achieved if the Nyquist pulse sn is sampled and digitized as illustrated in the graph of The horizontal and vertical illustrated grid lines may be defined by a scale of from 0 to 8 on each of the two coordinate axes I, Q. Arcs which may lie midway between two nominal arcs may be illustrated in The definition of the midway point between two nominal arcs as a limit radius is exemplary. For example, the respective limit radii may be shifted from the middle in either direction, as indicated by the dash-dot arcs in If the selection of the most probable nominal radius Rsi is made by the radius R and by the angle α, the range limits may no longer be circular arcs but may deform somewhat. In the vicinity of a symbol to be expected, S As an example, The polar coordinates Rs Referring to A quadrature mixer The output of the sampling device The generation of the auxiliary symbol Sh may be performed by an auxiliary-symbol generator Referring to The inputs of the controller Except for the differences described, the embodiment of The interface Although the present invention has been shown and described with respect to several preferred embodiments thereof, various changes, omissions and additions to the form and detail thereof, may be made therein, without departing from the spirit and scope of the invention. Patent Citations
Non-Patent Citations
Classifications
Legal Events
Rotate |